{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6VKT2P5EJ2MGHXYP373INKWMJZ","short_pith_number":"pith:6VKT2P5E","schema_version":"1.0","canonical_sha256":"f5553d3fa44e9863df0fdff686aacc4e6fa5b8220c34f270deb3e92f32fac06e","source":{"kind":"arxiv","id":"2503.07578","version":2},"attestation_state":"computed","paper":{"title":"Denoising Score Distillation: From Noisy Diffusion Pretraining to One-Step High-Quality Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Mingyuan Zhou, Oscar Leong, Tianyu Chen, Yasi Zhang, Ying Nian Wu, Zhendong Wang","submitted_at":"2025-03-10T17:44:46Z","abstract_excerpt":"Diffusion models have achieved remarkable success in generating high-resolution, realistic images across diverse natural distributions. However, their performance heavily relies on high-quality training data, making it challenging to learn meaningful distributions from corrupted samples. This limitation restricts their applicability in scientific domains where clean data is scarce or costly to obtain. In this work, we introduce denoising score distillation (DSD), a surprisingly effective and novel approach for training high-quality generative models from low-quality data. DSD first pretrains a"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2503.07578","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-10T17:44:46Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"93488c3a1d7a17cc48a9be718bfed855f3ce9cb8f45e1961e39664dd9e03a091","abstract_canon_sha256":"446dec2809664e15febe7c2d037e5e79074a98b4b82a7a0a255a2314ba0a2c91"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:55.435573Z","signature_b64":"K56uVg8OF7zHG4mkTxvLRZmdcm8aPKjLttt8vOWIDy6jXlANdBq+kde0+N5LyfGrMgSTh7BQZLq42ICdUur4Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f5553d3fa44e9863df0fdff686aacc4e6fa5b8220c34f270deb3e92f32fac06e","last_reissued_at":"2026-07-05T11:06:55.435028Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:55.435028Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Denoising Score Distillation: From Noisy Diffusion Pretraining to One-Step High-Quality Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Mingyuan Zhou, Oscar Leong, Tianyu Chen, Yasi Zhang, Ying Nian Wu, Zhendong Wang","submitted_at":"2025-03-10T17:44:46Z","abstract_excerpt":"Diffusion models have achieved remarkable success in generating high-resolution, realistic images across diverse natural distributions. However, their performance heavily relies on high-quality training data, making it challenging to learn meaningful distributions from corrupted samples. This limitation restricts their applicability in scientific domains where clean data is scarce or costly to obtain. In this work, we introduce denoising score distillation (DSD), a surprisingly effective and novel approach for training high-quality generative models from low-quality data. DSD first pretrains a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.07578","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2503.07578/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2503.07578","created_at":"2026-07-05T11:06:55.435088+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.07578v2","created_at":"2026-07-05T11:06:55.435088+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.07578","created_at":"2026-07-05T11:06:55.435088+00:00"},{"alias_kind":"pith_short_12","alias_value":"6VKT2P5EJ2MG","created_at":"2026-07-05T11:06:55.435088+00:00"},{"alias_kind":"pith_short_16","alias_value":"6VKT2P5EJ2MGHXYP","created_at":"2026-07-05T11:06:55.435088+00:00"},{"alias_kind":"pith_short_8","alias_value":"6VKT2P5E","created_at":"2026-07-05T11:06:55.435088+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12365","citing_title":"Ambient Diffusion Policy: Imitation Learning from Suboptimal Data in Robotics","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2512.10857","citing_title":"Generative Modeling from Black-box Corruptions via Self-Consistent Stochastic Interpolants","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ","json":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ.json","graph_json":"https://pith.science/api/pith-number/6VKT2P5EJ2MGHXYP373INKWMJZ/graph.json","events_json":"https://pith.science/api/pith-number/6VKT2P5EJ2MGHXYP373INKWMJZ/events.json","paper":"https://pith.science/paper/6VKT2P5E"},"agent_actions":{"view_html":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ","download_json":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ.json","view_paper":"https://pith.science/paper/6VKT2P5E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.07578&json=true","fetch_graph":"https://pith.science/api/pith-number/6VKT2P5EJ2MGHXYP373INKWMJZ/graph.json","fetch_events":"https://pith.science/api/pith-number/6VKT2P5EJ2MGHXYP373INKWMJZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ/action/storage_attestation","attest_author":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ/action/author_attestation","sign_citation":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ/action/citation_signature","submit_replication":"https://pith.science/pith/6VKT2P5EJ2MGHXYP373INKWMJZ/action/replication_record"}},"created_at":"2026-07-05T11:06:55.435088+00:00","updated_at":"2026-07-05T11:06:55.435088+00:00"}